Towards Interoperable, Cognitive and Autonomic IoT Systems: an Agent-based Approach. Claudio Savaglio, Giancarlo Fortino, Mengchu Zhou. IEEE WFIoT 2016

Towards Interoperable, Cognitive
and Autonomic IoT Systems:
an Agent-based Approach
Claudio Savaglio, Giancarlo Fortino, Mengchu Zhou
3rd IEEE World Forum on Internet of Things, Reston, VA, USA, 12 December 2016

HOW TO CITE THE MANUSCRIPT RELATED TO THESE SLIDES:
•
•
2

Outline
1. Background
2. Motivations and Proposal
3. Agent-based SO development
4. ACOSO middleware
5. Simulation
6. Conclusion and future work
3

Background – toward the Internet of Things (IoT)
4
“Intranets of Things”
are domain specific and
scarcely interoperable
[Z10].

Background – toward the Internet of Things (IoT)
5
“Intranets of Things”
are domain specific and
scarcely interoperable
[Z10].
Voluntary interoperability among
heterogeneous IoT platforms spanning
multiple application domains (INTER-
IoT H2020 EU research project
[WS16]).

Background – Smart Object (SO)
SOs are fundamental IoT building blocks [K10] -> SO-based IoT vision 6
Augmented real world objects autonomously interacting each others and
proactively providing cyber-physical services.

Motivations
Desiderata
Interoperability despite
functional/protocol heterogeneities
Autonomy
Awareness/Smartness
Dynamicity
Scalability
Reliability and Robustness
Price/Design Constraint
SO-based IoT systems development
poses notably challenges, both at thing
and at system level [G13].
Proposal: an agent oriented approach
(based on the ACOSO middleware) to
develop interoperable, self-steering and
heterogeneous SO-based IoT systems.
7

Agent-based SO development
input
output
A software Agent [L04] :
• is an autonomous, goal-directed entity;
• is situated in, is aware of, and reacts to its
environment;
• cooperates to accomplish its tasks.
Agents are able to:
• encapsulate complex functionalities and abstract heterogeneous resources;
• act as interoperability facilitators;
• fully support the development of complex, cooperative and adaptive
distributed systems [F12].
8

Agent-based SO development
Following the agent-based computing paradigm, (autonomic)
SOs are modeled as Agents and (cognitive) SO-based IoT
Systems as Multi agent Systems (M.A.S.).
Autonomic Computing [K03] Cognitive Networks [F09]
9

ACOSO (Agent-based COoperating SO) Middleware
ACOSO [F13] provides
• (in-the-small and in-the-large) SO programming model;
• platform neutral, event-driven and multi-layered architecture;
• Jade-based agent management and infrastructure of
communication.
Each ACOSO-based SO is a cooperating agent characterized by
its set of tasks, namely event-driven and state-based components
modeling agent behaviors and goals.
10

Communication
11
Main SO features
ACOSO Architecture
Smartness &
Knowledge
Service Provision
Sensing /
Actuation

ACOSO-based approach is
• fine-grained, since SOs can be designed with different degrees
of smartness;
• modular, because the ACOSO architecture consists of pluggable
components (e.g., device/communication adapters);
• efficient and flexible, with frozen spots (basic SO
functionalities already implemented) and hot spots (adapters,
new User Defined Tasks for specific SO service).
12

Simulation
Before the actual SO deployment phase, IoT Systems simulation
allows:
• investigating possible issues or unpredicted situations,
• testing IoT System design choices and setting parameters;
• validating of models, protocols and algorithms.
Currently IoT-specific simulators are not available yet [F16].
13

Goals:
• Simulating IoT System high-level communication among SOs
(modeled following the ACOSO approach) through OMNET++
[V10].
• Investigating how the SO population and its distribution may
influence the overall performances in the SO Discovery (SOD)
phase and in Information exchange (IE) phase.
Network Scale #SO #subnet Metrics Parameters
Small 50-100 1 (Discovery/RoundTrip)Time,
(Request/Message)DeliveryRatio
#SO,
TCP-based and
UDP-based protocols
Medium 250–500 5
Large 500-1000 10
14
Simulation

Simulations results:
1. When the SO number increases the performance worsens due
to wireless interferences and network congestion --> involve
as few SOs as possible.
2. TCP-based protocols guarantee best results in terms of
(Request/Message) Delivery Ratio but they imply higher values
of (Discovery/Roundtrip) Time --> avoid reliable protocols in
time-sensitive application scenarios.
3. Overlapping among multiple subnetworks increases network
congestion --> a well-thorough networks design and
deployment is recommended.
15
Simulation

Conclusion
IoT Systems need to be interoperable and self-steering.
The ACOSO-based approach represents a viable solution
i. to provide concrete interoperability among heterogeneous IoT
Systems at different levels of granularity;
ii. to enable autonomic and cognitive management mechanisms
that allow supporting the IoT Systems integration process.
Ongoing and Future Work
• To allow SOs virtualization by using edge- and cloud-based
infrastructures [F14].
• To pursue further application in specific domains (e.g., Internet
of Vehicles [C15]). 16

HOW TO CITE THE MANUSCRIPT RELATED TO THESE SLIDES:
•
•
17

THANK YOU FOR YOUR ATTENTION!
QUESTIONS?

References
• Zorzi, M. et al. "From today's intranet of things to a future internet of things: a wireless-and
mobility-related view." IEEE Wireless Communications 17.6 (2010): 44-51.
• Inter-IoT Website http://www.inter-iot-project.eu/ (last access December 2016)
• Kortuem, G. et al. "Smart objects as building blocks for the internet of things." IEEE Internet
Computing 14.1 (2010): 44-51.
• Gubbi, J. et al. "Internet of Things (IoT): A vision, architectural elements, and future directions."
Future Generation Computer Systems 29.7 (2013): 1645-1660.
• Luck, M., P. McBurney, and C. Preist. "A manifesto for agent technology: Towards next generation
computing." Autonomous Agents and Multi-Agent Systems 9.3 (2004): 203-252.
• Fortino, G., A. Guerrieri, and W. Russo. "Agent-oriented smart objects development." Computer
Supported Cooperative Work in Design (CSCWD), 2012 IEEE 16th Int. Conf. on. IEEE, 2012.
• Kephart, J. O., and D. Chess. "The vision of autonomic computing." Computer 36.1 (2003): 41-50.
• Fortuna, C., and M. Mohorcic. "Trends in the development of communication networks: Cognitive
networks." Computer networks 53.9 (2009): 1354-1376.
• Fortino, G., et al. "An agent-based middleware for cooperating smart objects." Int. Conf. on Practical
Applications of Agents and Multi-Agent Systems. Springer Berlin Heidelberg, 2013.
• Fortino, G., W. Russo, and C. Savaglio. "Simulation of Agent-oriented Internet of Things Systems."
Proc. 17th Workshop" From Objects to Agents. 2016.
• Varga, A. "OMNeT++." Modeling and Tools for Network Simulation. Springer Berlin Heidelberg, 2010.
35-59.
• Fortino, G. et al. "Integration of agent-based and cloud computing for the smart objects-oriented
iot." Computer Supported Cooperative Work in Design, Proc. of the IEEE 18th Int. Conf. on., 2014.
• Cheng, J. et al. "Routing in internet of vehicles: A review." IEEE Transactions on Intelligent
Transportation Systems 16.5 (2015): 2339-2352.

21
1 2
3 4
5
1
1
3
4
5
6
2
9
8
7
10
Small
Network
Medium
Network
Large
Network

Towards Interoperable, Cognitive and Autonomic IoT Systems: an Agent-based Approach. Claudio Savaglio, Giancarlo Fortino, Mengchu Zhou. IEEE WFIoT 2016

More Related Content

What's hot (19)

Similar to Towards Interoperable, Cognitive and Autonomic IoT Systems: an Agent-based Approach. Claudio Savaglio, Giancarlo Fortino, Mengchu Zhou. IEEE WFIoT 2016 (20)

Recently uploaded (20)

Towards Interoperable, Cognitive and Autonomic IoT Systems: an Agent-based Approach. Claudio Savaglio, Giancarlo Fortino, Mengchu Zhou. IEEE WFIoT 2016